The search for genes involved in the development of late-onset Alzheimer's disease (AD) has recently moved forward with the completion of several genome-wide scans, including one from the NIMH Genetics Initiative, a collaboration in which we were a partner. A large number of chromosomal regions have now been identified, though findings are often inconsistent between studies. This may be due to the heterogeneity of the families included as well as the etiologic complexity of late-onset AD. New strategies that reduce heterogeneity are needed to prioritize linked regions for fine mapping and identify genes for this disease. We intend to refine regions by defining more homogenous subgroups based on our extensive clinical and neuropathological phenotype information, using recently developed statistical genetics software appropriate for covariate analysis. In addition, we will improve our power by expanding our existing families through collection of an additional 100 family members including siblings, and all recently affected offspring of study probands, and their siblings. We will conduct fine mapping of regions prioritized through our phenotype analyses by increasing marker density with additional STR and SNP markers. We will then pursue plausible candidate genes in regions consistently showing linkage in our analyses. For these candidates, we will employ sibling pair linkage and family-based association analyses using multiple STR and SNP markers located within and around the candidate genes. In addition, we will conduct haplotype and single-locus case-control comparisons using a set of 100 unrelated cognitively healthy elderly controls from the same geographic region as our family probands. This analysis will provide supplementary evidence of associations observed in our familial set. This project provides a unique opportunity to evaluate a large sample of late-onset AD families with extensive phenotype information and available genome-wide marker genotypes, through emerging linkage and candidate-gene association analysis methods.